Seven plots were laid out for the 1985–6 growing season on the Chalk Downs near Lewes, East Sussex, to study the relationship between cultivation practice, soil conditions and erosion rates. Three plots were sown with cereals in autumn and three in spring. The seventh plot was ploughed and kept bare. In both spring and autumn, one plot was conventionally cultivated by ploughing and harrowing, one was direct drilled through burnt stubble and the third was direct drilled through the debris of the previous year's straw and stubble. Erosion was greatest from the conventionally cultivated plots and least from those direct drilled through unburnt stubble. Monitoring of moisture, infiltration, strength and compaction properties of soils suggests that from late autumn onwards soil conditions on the plots were very similar and the major factor determining the different erosion losses from the plots was the percentage vegetal cover (crop, weeds and residues). The results suggest that direct drilling through unburnt stubble may be a way of obtaining high yields with low erosion risk on the Downs, but the long-term consequences of such a practice remain unknown.

Peking ducklings were classified at 6 weeks of age according to their body weights in two categories: R, below the overall mean of the flock which were considered as retarded growth; and L, those of body weight equal to the overall mean and above which were considered as large. From each group, males and females were chosen to be parents of the next generation in a ratio 1:5 at 6 months of age and four systems of matings were conducted: RR, LL, RL and LR.

At all ages studied the large ducklings were significantly heavier than retarded growth ones and had significantly greater body gain, during the two successive generations.

The RL and LR matings were heavier than the RR and LL groups at the different ages studied, indicating an effect of heterosis and that the retarded growth character is controlled by few pairs of genes. The differences due to mating systems were not significant either in body weight or body gain, except in body weight at 18 and 24 weeks and relative gain at 6 weeks of age in which the differences were significant (P <0·05) in the second generation.

The other environmental factors studied were sex, month of hatch and 6-week body weight.

Heritability estimates were 0·489, 0·509, 0·576, 0·346, 0·713, 0·809, 0·651 and 0·386 for body weight at hatch 4, 6, 8, 12, 16, 18 and 24 weeks of age, respectively, using the data which were adjusted for the effects of sex, month of hatch, 6-week body weight and mating system. The same estimates were 0·482, 0·554, 0·578, 0·352, 0·713, 0·808, 0·650 and 0·386, respectively, using the same data but adjusted for month of hatch and sex only.

Phenotypic and genetic correlations between body weights at different ages for the two sets of data were estimated. The phenotypic correlations were negative between body weight at hatch and each of the body weights at 16 and 18 weeks of age for the two corrected sets of data. The genetic correlations were negative between hatch weight and each of the other body weights for the two adjusted sets of data and between body weight at 4 weeks of age and each of the body weights at 16, 18 and 24 weeks only for the first set of data. Phenotypic and genetic correlations between the body weights in older birds were highly positive.

Data on 841 purebred Bauscat, White Flander, Giza White and Baladi Red litters provided estimates of genetic, phenotypic and environmental parameters for gestation length, litter size at birth and at weaning, mortality and sex ratio at weaning. A total of 170 daughters (paternal half-sisters) of 76 sires were available for the analysis. No important differences were detected among breeds for litter traits studied except litter size at weaning (P < 0·001). Year of kindling affected (P < 0·001) gestation length, litter size at birth and mortality percentage. No clear patterns of the effect of parity and month of kindling on litter traits were observed. The sire of the doe and doe within sire affected most of the traits studied. Estimates of heritability indicated that the sire's genetic contribution to litter traits are much higher during the pre-natal period than for the suckling period. Estimates of repeatability for litter traits studied were relatively low. Genetic and phenotypic correlations between gestation length and other litter traits were relatively low. Litter size traits were positively correlated both genetically and phenotypically. Estimates of genetic and phenotypic correlation between litter size traits and preweaning mortality were generally negative. Phenotypic and environmental correlations were generally very similar in size and sign. Predicted direct selection was shown to give greater improvement in litter size at birth than indirect selection through litter size at weaning.

An experiment was carried out on a site where the copper content of herbage was relatively low (4·7–6·2 mg/kg d.m.) to examine the value of copper contained in soluble glass rumen boluses as a means of providing supplementary copper to calves.

Forty-five spring-born suckled calves (mean live weight 91 kg) were divided into three groups on the basis of body weight. Animals in group GB were given two boluses initially, those in group CI each received copper injections at approximately monthly intervals and those in group C remained as controls throughout. Group C calves became progressively hypocupraemic (minimum mean plasma copper 8·5 μmol/1) compared with those in groups GB and CI which showed an increase in plasma copper concentration. In addition measurements of plasma caeruloplasmin activity, blood haemoglobin concentration and erythrocyte superoxide dismutase activity all showed significant increases due to treatment. No response in live-weight gain was observed but overall live-weight gain was approximately 1·05 kg/day. Although the initial response in plasma copper was quicker with copper injections, measurements of erythrocyte superoxide dismutase activity and haemoglobin showed advantages of glass boluses not shared by copper injections. This possibly resulted from the slow release nature of copper in the glass. Under the conditions of the experiment where herbage molybdenum values were relatively low (1·2–1·9 mg/kg d.m.), copper contained in soluble glass boluses appeared to prevent hypocupraemia in calves as effectively as regular copper injections.

Periwinkle (Catharanthus roseus (L.) G. Don) is an important medicinal plant grown in different parts of India. Its leaves contain vincristine and vinblastin, alkaloids which are used in the treatment of various leukaemias and limphomas, and its roots contain ajmalcine, another alkaloid, used in the manufacture of antihypertensive drugs. Cultivation of periwinkle has recently gained importance in India. The yield of periwinkle can be increased with the use of mineral fertilizers (Pareek et al. 1981, 1982; Rajasekharan et al. 1982; Reghunath & Sethumadhavan, 1982a, b; Hegde, 1986). The effect of fertilizer application on the alkaloid concentration is not clearly established though there are a few reports which suggest that application of N increases alkaloid concentration in roots (Rajasekharan et al. 1982; Reghunath & Sethumadhavan, 1982a) and leaves (Pareek et al. 1982). Therefore, the present experiment was designed to study the effect of application of nitrogen and phosphatic fertilizers and farmyard manure on the root and leaf yields, and total alkaloid concentrations in roots and leaves in Catharanthus roseus (L.) G. Don.